Study
Optimizing colour for camouflage and visibility using deep learning: the effects of the environment and the observer's visual system
Fennell, J. G.; Talas, L.; Baddeley, R. J.; Cuthill, I. C.; Scott-Samuel, N. E.
Journal of The Royal Society Interface, 2019
Photographs of natural scenes were analysed to find the colour that would be hardest to detect against each background, computed separately for a trichromatic observer such as a human and for a dichromatic observer such as most hoofed mammals.
- Sample size
- natural scene images analysed computationally
The optimal concealing colour depends heavily on the observer’s visual system. Colours that stand out sharply to a trichromat can be close to optimally concealing against foliage for a dichromat, because the red–green distinction that makes them conspicuous is unavailable.
What it means
The authors' reading, and ours. Where they differ, that difference is the point.
How the authors put it
Camouflage must be assessed against the visual system of the animal that matters, not against human perception.
How NatureHQ reads it
This is the piece that makes the tiger make sense, and it is not a tiger study — which is exactly why the tiger version of the argument is so often stated badly. Deer and pigs, the animals a tiger eats, are dichromats and cannot separate orange from green. The tiger is not orange despite the forest being green; it is orange because to the animal it is hunting, orange and green are much the same colour. The striking appearance is a human-only problem.
- Computational rather than behavioural: no predator or prey was tested.
- Uses modelled visual systems rather than measurements from any individual prey species.
- Addresses colour, not the disruptive effect of the stripe pattern itself.
Asking what colour is invisible — to a deer
Is a colour that looks conspicuous to a human necessarily conspicuous to the animal it needs to hide from?
What this study is used for on NatureHQ
One study can inform several subjects. Here is everywhere this one is cited.
Published by: The Royal SocietyPublished by the Royal Society.
doi.org/10.1098/rsif.2019.0183
This reference was resolved automatically against Crossref on 2026-08-11.
NatureHQ summarises research in its own words and does not reproduce published text. Reviewed 2026-08-11.